Bayesian Estimation Tools

The GAUSS Bayesian Estimation Tools package provides a suite of tools for estimation and analysis of a number of pre-packaged models. The internal GAUSS Bayesian models provide quickly accessible, full-stage modeling including data generation, estimation, and post-estimation analysis. Modeling flexibility is provided through control structures for setting modeling parameters, such as burn-in periods, total iterations and others.


Posteriori Distribution of Lambda
Posteriori Distribution of Lambda

Posteriori Distribution of Sigma
Posteriori Distribution of Sigma

GAUSS Bayesian internal models

  • Univariate and multivariate linear models
  • Linear models with auto-correlated error terms
  • HB Interaction and HB mixture models
  • Probit models
  • Logit models
  • Dynamic two-factor model
  • SVAR models with sign restrictions

Individual modeling

Users can meet individual modelling needs by specifying key controls for the estimation algorithm including:

  • Number of saved iterations
  • Number of iterations to skip
  • Number of burn-in iterations
  • Total number of iterations
  • Inclusion of an intercept

Data loading and data generation

Users may load data into GAUSS for estimation and analysis using standard intrinsic GAUSS procedures. However, in addition, the Bayesian Analysis Module includes a data generation feature that allows users to specify true data parameters to build hypothetical data sets for analysis.

Easy to interpret stored results

The Bayesian application module stores all results in a single output structure. In addition the Bayesian module graphs draws of all parameters and the posterior distributions for all parameters.

  • Draws for all parameters at each iteration
  • Posterior mean for all parameters
  • Posterior standard deviation for all parameters
  • Predicted values
  • Residuals
  • Correlation matrix between Y and Yhat
  • PDF values and corresponding PDF grid for all posterior distributions
  • Log-likelihood value (when applicable)